Whine / squeal noises when machine idle

[bkraz]

I have been having very good luck using my "new" Bridgeport R2E3 mill. So far, everything seems to work perfectly, but I have one question that may also relate to CNC machines in general: When my machine is idle (axis drives enabled, but nothing is moving, spindle off), the machine emits a very high-pitched whistle noise (a squeal, or whine). The noise is constant and occurs at all positions of the axes. Disabling the axis drives stops the noise. When the machine is in motion the noise seems to lessen, but this may be just because the sound of the drive train overpowers it. The noise isn't really that annoying, I just want to be sure this isn't a sign of something failing. Also, I am using a 5HP rotary phase converter.

Any opinions on this? Thanks -Ben

[Al_The_Man]

This is sometimes apparent on DC pwm drives, especially older type that use velocity loop amplifiers, i.e. tach feedback type. At rest, especially if the gain is turned a little high.
The PWM signal is alternating between 0 and max plus Volts with the H bridge output turned on and off an equal amount in either direction so the motor is being 'caged' or held in position.
I do not find the same effect when switching over to a modern Torque (current) style amplifier.
IOW on the older amps like Westamp etc, this is normal.
Al.

[machintek]

The R2E# and the R2E4 both emitted a high pitched whine from the "chokes" that are part of the drive circuit. A lot of thes were installed in schools and they were hated for the noise. Bridgeport did come out with a set of chokes that were half of the original milli henry value but charged a lot of money for them. This could decrease the volume of noise but I never had anyone buy this.
The other possibility is that as mentioned above your gain is too high and the motors are going into an oscillation. In all probability it is the first choice above.

George

[NC Cams]

We found that the "harder" you want the servos to hold something stationary, the greater the possibility of the "hum/buzz" from the drive.

Reducing the gain did tend to reduce "hum/buzz" oscillation potential.

A good servo tune migh be in order and could do wonders....

[fredhh47]

We had the same problem with our R2E3, never managed to quiet it down. We kept the radio turned up and that helped drown it out, but you could never really shut out the noise, it would just get into you head and bore. Never heard about any replacement chokes, or any other remedy. That was one badly engineered machine. I could make a long list of the problems we had with it. Everybody breathed a sigh of relief when we shipped it out to be retrofitted with a Centroid control, which, BTW has worked almost flawlessly (and silently) for about 10 years. There was one very interesting quirk that we ran across with the Bridgeport Boss 9 software. As I recall, and this was a long time ago, if you used geometric functions in your program, and attempted to create a hyperbolic function where the "X" or "Y" value went to zero, the machine would stall and "dither" back and forth across the zero point until you hit e-stop. It was caused by the fact that the X or Y value approaching zero pushed the opposite value towards infinity, and the controller logic couldn't handle the value of infinity. We had a problem of this sort programming a mold for a foam plate, as the edge fluting was defined by just such a geometric function. The fluting was all around the edge, so whenever X or Y got to zero, the machine would have a hissy fit. We were able to solve the problem quite simply by rotating the entire program by .001 degree, so at no time did the X or Y value become EXACTLY zero. It only came up that one time. Apparently, nobody at Bridgeport ever considered that people would actually use the geometric functions in a real program, because they were clueless as to how to overcome the problem. We just happened on the solution by chance.

[NC Cams]

It never ceases to amaze me of the issues that programmers inadvertently cause.

I wonder if these things would happen if the programmers had to run the machines and make them do stuff BEFORE they shipped them.

A computer does not consider 1.99999999999999999999 to be equal to 2 but most humans do in a snap - for all intents and purposes it is EXCEPT when you have this inanimate object being so anal over a miniscule difference of 0.00000000000000000001 (I hope I counted the decimal places right!!!)